Summary
To be diagnosed is to be named. A diagnosis is a word placed between a person and the world — it reorganises the past, reframes the present, and constrains or expands the future. For autistic people, diagnosis can be profoundly liberating: finally, an explanation. It can also be profoundly limiting: now, a label. Often it is both at once.
This page does not argue for or against diagnosis. It examines what diagnosis is — philosophically, historically, and practically — and explores how the act of naming a person’s neurology shapes their life, their identity, and the systems around them. The aim is not neutrality for its own sake, but honesty: diagnosis is a powerful tool with real benefits and real costs, and anyone working in this field should understand both.
A brief history of the autism diagnosis
The word autism predates the condition it now names. Swiss psychiatrist Eugen Bleuler coined it in 1911, from the Greek autos (self), to describe a feature of schizophrenia — withdrawal into the self. The term migrated into developmental psychology in the 1920s, when Jean Piaget used it to describe pre-social states of child thought.
The condition we now call autism was first described clinically — though not by that name — by the Soviet child psychiatrist Grunya Sukhareva in 1926, who published detailed case descriptions of six boys whose presentation closely matches modern diagnostic criteria (Manouilenko & Bejerot, 2015). Her work, published in German, was largely overlooked for decades.
In 1943, Leo Kanner at Johns Hopkins described eleven children with “inborn autistic disturbances of affective contact,” emphasising social withdrawal and insistence on sameness. Independently — or perhaps not entirely independently, as Silberman (2015) has argued — Hans Asperger described similar children in Vienna in 1944, focusing on those with preserved language and intelligence. Asperger’s work remained largely unknown in the English-speaking world until Lorna Wing’s landmark paper in 1981 brought it to wider attention.
For decades, autism was classified under childhood schizophrenia. The “refrigerator mother” hypothesis — the psychoanalytic claim that autism was caused by cold, unloving parenting — dominated clinical thinking from the 1950s through the 1970s, causing immeasurable harm to families. This theory was eventually discredited by accumulating evidence of autism’s neurobiological basis, particularly Michael Rutter’s twin studies in the 1970s.
DSM-III (1980) first recognised autism as a distinct diagnostic category — “infantile autism” under “pervasive developmental disorders.” DSM-IV (1994) added Asperger’s disorder as a separate diagnosis, creating a formal distinction between “classic” autism and “high-functioning” variants. DSM-5 (2013) collapsed these back into a single category, “autism spectrum disorder,” with severity levels — a move that remains controversial. The ICD-11 (2019) largely followed suit, and notably included a reference to masking as a factor that clinicians should consider in assessment.
Each of these shifts — from schizophrenia to its own category, from single diagnosis to spectrum, from spectrum with subtypes back to unified spectrum — was not merely a technical adjustment. Each redrew the boundary of who counted as autistic, who could access services, and who was excluded.
What diagnosis does
It explains
For many autistic people, particularly those diagnosed in adulthood, the diagnosis is an explanation — not of what is wrong with them, but of why they have always felt different. The relief of late diagnosis is one of the most consistent themes in qualitative research on autistic adults’ experiences. Years of feeling wrong, of struggling in ways that others seemed not to, of being told to try harder — all of it suddenly has a name and a context. The diagnosis reframes a lifetime of experiences: not failing at being normal, but succeeding at being autistic in a world that wasn’t designed for you.
It provides access
In most countries, diagnosis is the gateway to services, accommodations, and legal protections. Without a formal diagnosis, an autistic person may be ineligible for educational support, workplace adjustments, disability benefits, or specialist services. The diagnosis is a bureaucratic key — and for many autistic people, particularly those with high support needs, it is essential to their wellbeing and survival.
This practical function should not be underestimated. Whatever philosophical reservations one might have about diagnostic categories, the material consequences of being diagnosed or undiagnosed are real.
It creates community
The autism diagnosis enables connection. It gives people a word to search for, a community to find, a shared language for experiences that may previously have felt incommunicable. Online autistic communities, self-advocacy organisations, and the neurodiversity movement itself are all made possible, in part, by the existence of a diagnostic category around which people can organise.
It shapes identity
Here the terrain becomes more complex. A diagnosis does not simply describe a pre-existing reality; it also shapes how a person understands themselves. The philosopher Ian Hacking (1995) described this as the “looping effect” of human kinds: when people are classified, they can become aware of their classification, and this awareness changes how they behave and how they understand themselves, which in turn can change the classification. Autism is a paradigm case. A person who receives an autism diagnosis may begin to notice, emphasise, or interpret their experiences through an autistic lens — not because they are fabricating symptoms, but because the diagnosis provides a framework that makes certain experiences salient in new ways.
This is not, in itself, a problem. All human beings make sense of themselves through available frameworks. But it does mean that diagnosis is not a neutral act of discovery. It is an act of interpretation that becomes part of the thing it interprets.
It shapes thinking about the person
A diagnosis also shapes how others think about the diagnosed person. Teachers, clinicians, employers, and family members may begin to see the person primarily through the lens of the diagnosis — attributing behaviours, preferences, and difficulties to “the autism” rather than to the whole person in context. This is sometimes called diagnostic overshadowing: when a diagnosis becomes so prominent that other factors (physical health, trauma, environment, personality) are overlooked.
For autistic people with intellectual disability, diagnostic overshadowing is a particular risk. Pain, depression, grief, and discomfort may be dismissed as “just part of the autism” rather than investigated and addressed.
What diagnosis costs
It pathologises
The DSM-5 defines autism spectrum disorder in the language of deficit: “persistent deficits in social communication and social interaction” and “restricted, repetitive patterns of behavior, interests, or activities.” To be diagnosed, a person must meet criteria framed entirely in terms of what they lack or do wrongly compared to a neurotypical norm. There is no criterion for autistic strengths, no acknowledgement of sensory richness, intense focus, or pattern recognition. The diagnostic act, as currently constituted, requires the clinician to document a person’s inadequacies.
This is not a minor stylistic point. The language of diagnosis enters reports, files, and records. It shapes how the person is discussed in case conferences, how their progress is measured, and what goals are set for them. When the diagnostic language is exclusively deficit-based, it tilts the entire system of support towards remediation rather than accommodation.
It can harm through delay or denial
The benefits of diagnosis depend on actually receiving one. Diagnostic bias — by gender, ethnicity, class, age, and presentation style — means that many autistic people are diagnosed late or not at all. Autistic women and girls are diagnosed later than men and boys, in part because of camouflaging (see: masking-and-camouflaging) and in part because diagnostic instruments were developed and validated primarily on male samples. Autistic people from ethnic minorities and lower socioeconomic backgrounds face additional barriers. And autistic adults who present with co-occurring mental health conditions are frequently misdiagnosed — with personality disorders, anxiety, depression, or eating disorders — before their autism is recognised.
Late diagnosis is associated with worse mental health outcomes, not because of the autism itself but because of years of unsupported struggle, misdiagnosis, and self-blame. The diagnosis, when it finally arrives, can be both a relief and a grief: relief at understanding, grief at what might have been different.
It fixes what is fluid
Diagnostic categories create the impression of stable, bounded entities. But autism — like all neurodevelopmental conditions — is not a fixed state. Autistic people change over the lifespan. Their presentation varies across contexts. Some children who meet diagnostic criteria at age three no longer meet them at age ten — though whether this represents genuine change, compensation, or the limitations of the diagnostic instruments is debated. The spectrum is not a line from “mild” to “severe” but a multidimensional space that each person moves through differently.
The DSM-5’s severity levels (Level 1: “requiring support”; Level 2: “requiring substantial support”; Level 3: “requiring very substantial support”) attempt to capture this variation but are widely criticised for being vague, context-dependent, and easily misunderstood as permanent rankings rather than dynamic descriptions of current need.
It draws arbitrary lines
Every diagnostic system must decide where “autistic” ends and “not autistic” begins. This boundary is inevitably somewhat arbitrary. The person who scores one point below the cut-off on the ADOS-2 has a very similar neurological profile to the person who scores one point above it, but only one of them receives the diagnosis — and with it, the access, the identity, the community, and the label.
The question of whether self-identification without formal diagnosis is valid is one of the most contested in contemporary autism discourse. Neurodiversity advocates often support self-identification, arguing that diagnostic barriers disproportionately exclude women, people of colour, and those who cannot afford private assessment. Critics worry that overinclusion dilutes the category and redirects resources away from those with the highest support needs. Both concerns have merit.
The ontological question: What is autism?
Beneath the practical debates lies a deeper philosophical question: what kind of thing is autism? Is it a natural kind — a real entity in the world that diagnosis discovers? Or is it a construct — a useful but ultimately arbitrary way of dividing up the continuous variation of human neurology?
The philosopher Robert Chapman (2020) has argued that autism is best understood not as a natural kind or a social construction but as what Iris Marion Young called a “serial collective” — a group defined not by shared essence but by shared external circumstances. On this view, autistic people are united not by having the same brain but by being similarly affected by the same social, material, and environmental structures. This allows autism to be real and politically meaningful without requiring it to be biologically uniform — a useful resolution given the enormous genetic and phenotypic heterogeneity that characterises the condition.
The philosopher of psychiatry Monique Botha (2021) has applied critical realist philosophy to argue that autism has its own ontology — it exists independently of our theories about it — but that our theories are fallible and shaped by the social contexts in which they are produced. This means that medicalised accounts of autism can be critiqued not because autism isn’t real, but because the way we describe it is necessarily partial and influenced by power, perspective, and historical context.
These are not merely academic questions. How we answer them determines whether we think self-diagnosis is valid, whether the diagnostic criteria need revision, whether autism is a “thing” that can be cured, and whether the enormous heterogeneity within the category represents a problem to be solved or a feature to be expected.
Diagnosis and this wiki
This wiki does not require a formal diagnosis for a person’s experience to count. It takes the position that:
- Diagnosis is a tool, not a truth. It can be immensely useful, but it is not the only way to understand oneself or to access support.
- The current diagnostic criteria are deficit-framed and biased, and this should be acknowledged by anyone who uses them.
- Diagnostic overshadowing is a real risk, especially for autistic people with intellectual disability. A diagnosis should open doors, not close them.
- The experience of sensory processing differences does not depend on having a diagnostic label. A person who is overwhelmed by fluorescent lighting is overwhelmed regardless of whether they have a piece of paper that says so.
- The voices of autistic people — diagnosed, self-identified, and seeking answers — are all worth listening to.
Open questions
- What would a non-deficit-based diagnostic framework for autism look like? Could it be done without losing the practical utility of diagnosis for accessing services?
- How should diagnostic systems account for masking and camouflaging, which systematically lead to underdiagnosis in women, non-binary people, and those without intellectual disability?
- How can diagnostic equity be improved — particularly for people from ethnic minorities, lower socioeconomic backgrounds, and cultures in which Western psychiatric categories may not map neatly onto local understandings?
- Is the unified “autism spectrum disorder” of the DSM-5 the best way to organise this category, or would a return to subtypes (based on, e.g., sensory processing profiles, genetic architecture, or support needs) serve people better?
- What is the relationship between formal diagnosis and self-identification? Can they coexist productively, or are they in tension?
Key sources
- Kanner, L. (1943). Autistic disturbances of affective contact. Nervous Child, 2, 217–250.
- Asperger, H. (1944). Die “autistischen Psychopathen” im Kindesalter. Archiv für Psychiatrie und Nervenkrankheiten, 117, 76–136.
- Manouilenko, I. & Bejerot, S. (2015). Sukhareva — prior to Asperger and Kanner. Nordic Journal of Psychiatry, 69(6), 479–482. doi: 10.3109/08039488.2015.1005022
- Wing, L. (1981). Asperger’s syndrome: a clinical account. Psychological Medicine, 11(1), 115–129.
- Volkmar, F.R. & McPartland, J.C. (2014). From Kanner to DSM-5: autism as an evolving diagnostic concept. Annual Review of Clinical Psychology, 10, 193–212. doi: 10.1146/annurev-clinpsy-032813-153710
- Volkmar, F.R. et al. (2021). The diagnosis of autism: from Kanner to DSM-III to DSM-5 and beyond. Journal of Autism and Developmental Disorders, 52, 4098–4119. doi: 10.1007/s10803-021-04904-1
- Silberman, S. (2015). NeuroTribes: The Legacy of Autism and How to Think Smarter About People Who Think Differently. London: Allen & Unwin.
- Hacking, I. (1995). The looping effects of human kinds. In D. Sperber, D. Premack & A.J. Premack (Eds.), Causal Cognition: A Multidisciplinary Debate (pp. 351–394). Oxford: Clarendon Press.
- Chapman, R. (2020). The reality of autism: on the metaphysics of disorder and diversity. Philosophical Psychology, 33(6), 799–819. doi: 10.1080/09515089.2020.1751103
- Botha, M. (2021). A critical realist approach on autism: ontological and epistemological implications for knowledge production in autism research. Frontiers in Psychology, 12, 713565. doi: 10.3389/fpsyg.2021.713565
- Botha, M. & Cage, E. (2022). Autism research is in crisis: a mixed method study of researchers’ constructions of autistic people and autism research. Frontiers in Psychology, 13, 1050897.
- Runswick-Cole, K., Mallett, R. & Timimi, S. (Eds.) (2016). Re-thinking Autism: Diagnosis, Identity and Equality. London: Jessica Kingsley Publishers.
- American Psychiatric Association (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Washington, DC: APA.
- World Health Organisation (2019). International Classification of Diseases (11th revision). Geneva: WHO.
- Russell, G. (2020). Critiques of the neurodiversity movement. In S. Kapp (Ed.), Autistic Community and the Neurodiversity Movement (pp. 287–303). Singapore: Springer.